Task

PFMBench PDB-BIND: PDBbind

PDBbind. Scored with Spearman on PDBbind. Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1.

12 evaluations · 12 metric rows

Overview

PDBbind. Scored with Spearman on PDBbind. Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluation design

Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.

Benchmarks

PFMBench

These source-backed links do not make different protocols or scores interchangeable.

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Run instructions

No runnable recipe has been reviewed for this task. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

A task describes a biological question. Choose a linked protocol to obtain concrete split and scoring instructions.

Published comparisons

Explore the results reported under one evaluation protocol. Each figure keeps its source, dataset and metric together; it is not a ranking across studies. The pooled view gathers every source table that reports the same metric and names what it does not hold constant.

Comparison and metric

PFMBench PDB-BIND: PDBbind

spearman (correlation) · Higher values are better for this metric.

Every method PFMBench reports on PDBbind, scored with Spearman on PDBbind.

Evaluation protocol · PDBbind (PFMBench split)

ESM-2 · Configuration · Author-reported evaluation0.14677
VenusPLM · Configuration · Author-reported evaluation0.16536
ESM-C · Configuration · Author-reported evaluation0.14692
ProtGPT2 · Configuration · Author-reported evaluation0.13503
PGLM · Configuration · Author-reported evaluation0.16877
ProtT5 · Configuration · Author-reported evaluation0.20105
DPLM · Configuration · Author-reported evaluation0.13659
SaProt · Configuration · Author-reported evaluation0.15549
ProstT5 · Configuration · Author-reported evaluation0.18344
ProtST · Configuration · Author-reported evaluation0 .19514
ESM3 · Configuration · Author-reported evaluation0.15572
ProTrek · Configuration · Author-reported evaluation0.17322

Source order is preserved. Plotted marks show point estimates; uncertainty, where reported, is retained in the printed values and table. Differences do not establish statistical significance.

pfmbench primary benchmark evidence · Table 1, row(PDBbind)

Values, uncertainty and evidence

spearman: original source values
Tested entity	Printed value	Uncertainty	Evidence
ESM-2 · Configuration	0.14677 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ESM-2 [ 35 ]), column(PDBBind)
VenusPLM · Configuration	0.16536 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(VenusPLM [ 59 ]), column(PDBBind)
ESM-C · Configuration	0.14692 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ESM-C), column(PDBBind)
ProtGPT2 · Configuration	0.13503 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ProtGPT2 [ 12 ]), column(PDBBind)
PGLM · Configuration	0.16877 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(PGLM [ 5 ]), column(PDBBind)
ProtT5 · Configuration	0.20105 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ProtT5 [ 11 ]), column(PDBBind)
DPLM · Configuration	0.13659 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(DPLM [ 68 ]), column(PDBBind)
SaProt · Configuration	0.15549 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(SaProt [ 55 ]), column(PDBBind)
ProstT5 · Configuration	0.18344 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ProstT5 [ 21 ]), column(PDBBind)
ProtST · Configuration	0 .19514 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ProtST [ 71 ]), column(PDBBind)
ESM3 · Configuration	0.15572 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ESM3 [ 18 ]), column(PDBBind)
ProTrek · Configuration	0.17322 correlation	Not reported	Author-reported evaluation · source checkedpfmbench primary benchmark evidence · Table 3, row(ProTrek [ 56 ]), column(PDBBind)

Scope and limitations

Author-reported numbers, source checked but not independently reproduced.
The metric differs by task, taken from Table 1, so these figures cannot be averaged into one score.
Table 3 scores come from adapter fine-tuning; the ProteinGym figure is zero-shot and is not comparable to them.

Source transcription and grouping reviewed by automated source review on 2026-09-18. These experiments were not independently reproduced by rewire.

Tested entities and results

Release 2026-09-17-134cd1815de8 · 12 evaluations · 12 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.

Evaluation setupMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
DPLM on PFMBench PDB-BIND: PDBbind Configuration: DPLMTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.13659 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(DPLM [ 68 ]), column(PDBBind) Source checking is not independent reproduction.
ESM-2 on PFMBench PDB-BIND: PDBbind Configuration: ESM-2Task: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.14677 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ESM-2 [ 35 ]), column(PDBBind) Source checking is not independent reproduction.
ESM-C on PFMBench PDB-BIND: PDBbind Configuration: ESM-CTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.14692 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ESM-C), column(PDBBind) Source checking is not independent reproduction.
ESM3 on PFMBench PDB-BIND: PDBbind Configuration: ESM3Task: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.15572 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ESM3 [ 18 ]), column(PDBBind) Source checking is not independent reproduction.
PGLM on PFMBench PDB-BIND: PDBbind Configuration: PGLMTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.16877 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(PGLM [ 5 ]), column(PDBBind) Source checking is not independent reproduction.
ProstT5 on PFMBench PDB-BIND: PDBbind Configuration: ProstT5Task: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.18344 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ProstT5 [ 21 ]), column(PDBBind) Source checking is not independent reproduction.
ProtGPT2 on PFMBench PDB-BIND: PDBbind Configuration: ProtGPT2Task: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.13503 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ProtGPT2 [ 12 ]), column(PDBBind) Source checking is not independent reproduction.
ProTrek on PFMBench PDB-BIND: PDBbind Configuration: ProTrekTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.17322 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ProTrek [ 56 ]), column(PDBBind) Source checking is not independent reproduction.
ProtST on PFMBench PDB-BIND: PDBbind Configuration: ProtSTTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0 .19514 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ProtST [ 71 ]), column(PDBBind) Source checking is not independent reproduction.
ProtT5 on PFMBench PDB-BIND: PDBbind Configuration: ProtT5Task: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.20105 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(ProtT5 [ 11 ]), column(PDBBind) Source checking is not independent reproduction.
SaProt on PFMBench PDB-BIND: PDBbind Configuration: SaProtTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.15549 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(SaProt [ 55 ]), column(PDBBind) Source checking is not independent reproduction.
VenusPLM on PFMBench PDB-BIND: PDBbind Configuration: VenusPLMTask: PFMBench PDB-BIND: PDBbindDataset subset: PDBbind (PFMBench split) Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Author-reported evaluation · Evaluation metadata: source checked
0.16536 spearman Unit: correlation · Direction: higher	Uncertainty: Not reported Scored: Not reported · Eligible: Not reported	source checkedpfmbench primary benchmark evidence · Table 3, row(VenusPLM [ 59 ]), column(PDBBind) Source checking is not independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

1 evidence row matching the loaded filters

Claims, original sources and review scope · Release 2026-09-17-134cd1815de8
Property and statement	Original source and location	Review and provenance
Relationship: part of discovery-benchmark-pfmbench Individual claims	pfmbench primary benchmark evidence Original source ↗ Table 1, row(PDBbind) Version: 2506.14796v1 Retrieved: 2026-09-16T21:06:29.716004+00:00	source checked automated source review · 2026-09-18 Audit details Primary-source transcription with no human sign-off and no independent reproduction. Field: `links:part_of:discovery-benchmark-pfmbench` Claim: pfmbench-association-pdb-bind Source artifact SHA-256: `59c7bbb888e8e91f33c1e2cabfde062c32381d6ca727d23b6655c977aabf97a2` Hash scope: Exact retrieved primary paper artifact bytes. Inspected artifact

All evidence (CSV)All evidence (JSONL)How to read the evidence

Sources and history

Release 2026-09-17-134cd1815de8 · Record review: source checked

1 source records and release history

pfmbench primary benchmark evidence · Original source · 2506.14796v1

Download this release

Technical metadata and extraction receipts

Stable ID: pfmbench-task-pdb-bind

areas: proteins-complexes
tasks: PDBbind
metric: Spearman
metric direction: higher
dataset: PDBbind
protocol: Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1.
source locator: Table 1, row(PDBbind)
comparison panels: id: pfmbench-panel-pdb-bind; title: PFMBench PDB-BIND: PDBbind; protocol id: pfmbench-task-pdb-bind; dataset id: pfmbench-dataset-pdbbind; metric: spearman; unit: correlation; direction: higher; result ids: pfmbench-result-esm-2-pdb-bind-spearman; pfmbench-result-venusplm-pdb-bind-spearman; pfmbench-result-esm-c-pdb-bind-spearman; pfmbench-result-protgpt2-pdb-bind-spearman; pfmbench-result-pglm-pdb-bind-spearman; pfmbench-result-prott5-pdb-bind-spearman; pfmbench-result-dplm-pdb-bind-spearman; pfmbench-result-saprot-pdb-bind-spearman; pfmbench-result-prostt5-pdb-bind-spearman; pfmbench-result-protst-pdb-bind-spearman; pfmbench-result-esm3-pdb-bind-spearman; pfmbench-result-protrek-pdb-bind-spearman; source ids: expansion-p3-pfmbench; source locator: Table 1, row(PDBbind); context: Every method PFMBench reports on PDBbind, scored with Spearman on PDBbind.; caveats: Author-reported numbers, source checked but not independently reproduced.; The metric differs by task, taken from Table 1, so these figures cannot be averaged into one score.; Table 3 scores come from adapter fine-tuning; the ProteinGym figure is zero-shot and is not comparable to them.; review: method: automated_source_review; date: 2026-09-18

Related records

part of: PFMBench
subject: PFMBench PDB-BIND: part of discovery-benchmark-pfmbench
benchmark: DPLM on PFMBench PDB-BIND: PDBbind
benchmark: ESM-2 on PFMBench PDB-BIND: PDBbind
benchmark: ESM-C on PFMBench PDB-BIND: PDBbind
benchmark: ESM3 on PFMBench PDB-BIND: PDBbind
benchmark: PGLM on PFMBench PDB-BIND: PDBbind
benchmark: ProstT5 on PFMBench PDB-BIND: PDBbind
benchmark: ProtGPT2 on PFMBench PDB-BIND: PDBbind
benchmark: ProTrek on PFMBench PDB-BIND: PDBbind
benchmark: ProtST on PFMBench PDB-BIND: PDBbind
benchmark: ProtT5 on PFMBench PDB-BIND: PDBbind
benchmark: SaProt on PFMBench PDB-BIND: PDBbind
benchmark: VenusPLM on PFMBench PDB-BIND: PDBbind

Suggest a correction